Recovery of Varicella-Zoster Virus Specific T Cell Immunity after T Cell Depleted Allogeneic Transplantation Requires Symptomatic Virus Reactivation

Transcription

1 Recovery of Varicella-Zoster Virus Specific T Cell Immunity after T Cell Depleted Allogeneic Transplantation Requires Symptomatic Virus Reactivation Eva Distler, 1 Elke Schnürer, 1 Eva Wagner, 1 Charis von Auer, 1 Bodo Plachter, 2 Daniela Wehler, 1 Christoph Huber, 1 Karin Kolbe, 1 Ralf Georg Meyer, 1 Wolfgang Herr 1 Reactivated varicella-zoster virus (VZV) infection causes herpes zoster and commonly occurs after allogeneic hematopoietic stem cell transplantation (allo-hsct). Because VZV-specific T cell immunity is essential to prevent virus reactivation, we developed an interferon-g enzyme-linked immunosorbent spot (ELISPOT) assay for the sensitive detection of VZV-reactive T cells at the single-cell level ex vivo. We used this assay to monitor the frequency of VZV-reactive T cells in 17 seropositive patients during the first year after T cell depleted allo-hsct. The patients did not receive anti-herpesvirus prophylaxis after stem cell engraftment. Independent of the magnitude of transferred donor immunity, VZV-reactive T cell numbers decreased to low levels (median, 2/mL; range, to 35/mL) in peripheral blood early after transplantation. Only patients with subsequent zoster (n 5 5) exhibited a dramatic boost in VZV-reactive T cells (median, 366/mL; range, 158 to 756/mL), which was induced by the reactivation event. The postzoster VZV-reactive T cell levels were similar to those seen in healthy virus carriers. In contrast, antiviral T cell levels remained low in patients without VZV disease. Our results demonstrate that VZV-specific T cell immunity recovered efficiently during zoster in T cell depleted allo-hsct recipients. It did not reconstitute spontaneously in nonzoster patients, even in the absence of antiviral prophylaxis. Prospective studies should investigate whether VZV vaccination can substitute for natural resensitization by virus disease. Biol Blood Marrow Transplant 14: (28) Ó 28 American Society for Blood and Marrow Transplantation KEY WORDS: Herpes zoster, Varicella-zoster virus, ELISPOT, Interferon-g, T cell depletion INTRODUCTION 1 Department of Medicine III, Hematology and Oncology, and 2 Institute of Virology, Johannes Gutenberg-University, Mainz, Germany. R.G.M. and W.H. share senior authorship of this article. This work was supported by Deutsche Forschungsgemeinschaft grants SFB432/A13 (to W.H.), KFO183/TP5 (to W.H.), and KFO183/TP8 (to B.P.). Financial disclosure: See acknowledgments on page Correspondence and reprint requests: Wolfgang Herr, MD, Department of Medicine III, Hematology and Oncology, Johannes Gutenberg-University of Mainz, Langenbeckstr 1, 5511 Mainz, Germany ( w.herr@3-med.klinik.uni-mainz.de). Received May 21, 28; accepted September 5, /8/1412-1$34./ doi:1.116/j.bbmt Primary infection with varicella-zoster virus (VZV) causes chickenpox in nearly all humans [1]. Subsequently, the virus establishes life-long latency in neurons of the cranial and spinal ganglia. VZV can reactivate from this reservoir and produce herpes zoster, which is associated with pain, scarring, and postherpetic neuralgia. If the virus disseminates to the abdominal cavity or brain, then the infection may be life-threatening. Symptomatic VZV reactivation generally results from a decline in virus-specific cell-mediated immunity, a consequence of immunosenescence or of immunosuppressive disease or therapy [2]. The risk of reactivated disease is greatest in patients who have undergone allogeneic hematopoietic stem cell transplantation (allo-hsct), particularly in those treated with T cell depleting agents [3,4]; however, herpes zoster also occurs in T cell replete HSCT patients, with an incidence of 25% to 4% [5,6]. Although acyclovir prophylaxis effectively reduces VZV reactivation, it does not provide complete protection, and disease may appear several years posttransplantation [7-1]. VZV-specific T cell immunity appears to be essential to the prevention and control of VZV reactivation [2]. This immunity traditionally has been measured in HSCT patients using proliferation assays [7,9,11-13]. 1417

2 1418 E. Distler et al. Biol Blood Marrow Transplant 14: , 28 But although these assays can detect cell-mediated immune responses to VZV, they are unable to assess the precursor frequency of antiviral T cells. Consequently, we established an interferon (IFN)-g enzyme-linked immunosorbent spot (ELISPOT) assay that allows the sensitive detection of VZV-reactive T cells at the single-cell level in peripheral blood mononuclear cells (PBMCs) ex vivo. We used this assay to monitor the frequency of antiviral T cells in 17 seropositive patients during the first year after T cell depleted allo- HSCT. Our results show that VZV-reactive T cells decreased to very low levels early post-hsct. These cells in vivo expansion to levels detected in healthy virus carriers was observed only after symptomatic VZV disease. METHODS Patients The study protocol was designed to prospectively analyze immune reactivity to VZV in patients with hematologic malignancies after undergoing T cell depleted reduced-intensity allo-hsct. The pretransplantation conditioning regimen comprised fludarabine (Flu), melphalan (Mel), and the lymphodepleting antibody alemtuzumab [14]. All of the study patients had a history of chickenpox before transplantation and were VZV-seropositive. VZV serostatus also was positive for stem cell donors of sibling origin, but was unavailable for unrelated stem cell donors. It was determined by quantitative IgG and qualitative IgM enzyme-linked immunosorbent assay (Virion- Serion, Würzburg, Germany). All patients received herpesvirus prophylaxis with oral famciclovir (25 mg 3 times daily) until day 13 post-hsct, but not thereafter. VZV disease was diagnosed clinically and was confirmed by VZV polymerase chain reaction. It responded well to 1-week treatment with intravenous aciclovir (1 mg/kg body weight 3 times daily), followed by oral famciclovir (25 mg 3 times daily) over 3 to 4 weeks. The study was approved by the local ethics committee and was performed in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants. PBMCs were collected from study patients every 1 to 2 months throughout the first year after allo-hsct. The 1-year incidence of VZV disease was 35% for all patients treated in this protocol. The study population included 5 patients who developed herpes zoster during the first year after transplantation and 12 patients with no clinical signs of VZV reactivation during this period. The 2 groups did not differ significantly (P..5; Student s t- test) with regard to median age (57 vs 55 years) and median numbers of circulating CD4 1 T cells (35 vs 34/mL), CD8 1 T cells (5 vs 6/mL), and B cells (8 vs 3/ ml) measured at day 15 post-hsct before onset of VZV disease. On the same day, both groups had similar percentages of mixed donor T cell chimeras (2% vs 33%). Patients 1, 5, 6, 7, 1, 11, and 14 received prophylactic CD8-depleted donor lymphocyte infusions (DLIs) [14]. None of the patients developed acute graft-versus-host disease (agvhd) grade II-IV or extensive chronic GVHD (cgvhd) throughout the observation period. None received prednisone therapy. All patients were treated in the same study protocol, which scheduled the start of cyclosporine A (CsA) taper on day 135 after related donor transplantations and on day 175 after unrelated donor transplantations [14]. The zoster and nonzoster groups did not differ in terms of the proportion of patients requiring prolonged CsA treatment because of persistent moderate GVHD. Donor PBMCs analyzed for anti-vzv reactivity were consistently derived from redundant pilot tubes of leukapheresis products and were available from the stem cell donors of patients 1 to 11, 13, and 15. The respective leukaphereses were performed without preceding granulocyte-colony stimulating factor (G-CSF) treatment in the donors of patients 1, 5, 6, 7, 1, and 11 to prepare CD8-depleted DLIs [14]. In contrast, leukapheresis products of the donors for patients 2, 3, 4, 8, 9, 13, and 15 were harvested after G-CSF mobilization, because they were used for allo-hsct. The proportion of donors without G-CSF treatment was 4% and 5% in the analyzed donor cohorts of zoster and nonzoster patients, respectively. Because of this nearly equal distribution, the potential immunosuppressive effects induced by G-CSF were not expected to differ significantly between both donor cohorts. IFN-g ELISPOT Assay IFN-g ELISPOT assays were performed with cryopreserved PBMCs as described previously [15]. The freezing and thawing procedures were conducted according to standardized protocols for all PBMC samples. ELISPOT responders were total PBMCs (5 1 5 /well) or CD4 1 and CD8 1 T cells (1 1 5 / well) selected from PBMCs using CD4/CD8 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). Purified T cells were plated with /well autologous monocyte-derived dendritic cells [15]. Cells were incubated over 4 hours with lysates [15] prepared from VZV-infected and uninfected Vero cells (each 25 mg/ml), respectively. Infection of Vero cells with the VZV ROD strain was performed over 7 to 9 days before lysate production (Advanced Biotechnologies, Columbia, MD). Positive control wells contained PBMCs stimulated with 5 mg/ml phytohemagglutinin (PHA; Murex Pharmaceuticals, Dartford, UK). To determine the HLA restriction of VZVreactive T cells, 1 mg/ml of IgG2a monoclonal antibodies (mab) for blocking pan-hla class I (W6/32)

3 Biol Blood Marrow Transplant 14: , 28 Zoster drives posttransplant recovery of VZV-specific T cells 1419 or HLA-DR (L243) were added to the assay medium. Irrelevant IgG2a isotypes were included as controls. Antibodies were concentrated from supernatants of hybridomas purchased from American Type Culture Collection (Manassas, VA). Spot-forming cell (SFC) values in the diagrams are means of duplicate or triplicate wells. The frequency of VZV-reactive T cells was calculated based on SFCs per PBMCs of leukapheresis products, or SFCs per ml peripheral blood. The latter was derived from the SFC count per plated PBMCs and the related PBMC count per ml of peripheral blood. The PBMC count was assessed by measuring the percentage of total lymphocytes and monocytes in white blood cell differential analysis. Student s t-test was used to determine the significance of the results. Proliferation Assay Cryopreserved PBMCs were thawed and incubated at 1e5 cells/well for 6 days with VZV-infected Vero cell lysate or uninfected control lysate (1 mg/ml each) in 96-well round-bottomed microtiter plates in a total volume of 2 ml of AIM-V medium (GIBCO-BRL, Grand Island, NY) supplemented with 1% human serum. Wells with medium instead of lysates or with PHA (5 mg/ml) were used as controls. 3 H-thymidine (.5 mci/well; Amersham Life Science, Braunschweig, Germany) was added 18 hours before assay termination, and incorporation was measured with a TopCount scintillation counter (Wallace, Turku, Finland). The results represent means of triplicates with standard deviations indicated. RESULTS AND DISCUSSION Frequency Analysis of VZV-Reactive T Cells in Healthy Virus Carriers We first established an IFN-g ELISPOT assay to determine the precursor frequency of VZV-reactive T cells in PBMCs ex vivo. Given the comparatively low number of molecularly defined T cell epitopes of VZV, we used a VZV-infected cell lysate and an uninfected control lysate as the antigen sources. Figure 1A shows that the virus lysate was specifically recognized by T cells of VZV-seropositive healthy volunteers in a concentration-dependent manner. Because maximum reactivity was observed at 25 mg/ml in most donors, we chose this lysate concentration for all subsequent assays. Further experiments with anti-hla blocking antibodies and T cell subpopulations demonstrated that the anti-vzv memory response was dominated by HLA-class II restricted CD4 1 T cells, whereas VZV-reactive CD8 1 T cells were not detected (Figure 1B and C). This finding confirmed the findings of previous studies that were unable to identify VZVspecific CD8 1 T cells in immunocompetent donors by ex vivo analysis, even when oligopeptides and live viral vaccine were used for stimulation instead of lysates [16,17]. We believe that low numbers of circulating virus-specific CD8 1 T cells contribute to viral persistence, and possibly may reflect immune escape mechanisms, such as VZV-induced down-regulation of HLA-class I [18]. We next validated the IFN-g ELISPOT assays in 15 VZV-seropositive healthy volunteers who had a median frequency of 444 (range, 15 to 1134) VZV-reactive T cells/ml of peripheral blood (Figure 2). None of them recognized the uninfected control lysate (data not shown). In contrast, VZV-reactive IFN-g spot production was not observed in 5 seronegative healthy donors (Figure 2). Altogether, these results demonstrate the high specificity and sensitivity of the ELISPOT assay. Posttransplantation Recovery of VZV-Specific T Cell Immunity after VZV Disease We used the IFN-g ELISPOT assay to monitor the frequency of VZV-reactive T cells in 5 seropositive patients who developed localized herpes zoster 2 to 7 months after undergoing allo-hsct (Figure 3). The pretransplantation preparative regimen of these patients included T cell depletion with alemtuzumab and reduced-intensity conditioning (RIC) with Flu and Mel [14]. Immediately before the onset of clinical symptoms, the number of circulating antiviral T cells was extremely low (median, 2/mL; range, to 35/ ml). However, these counts increased dramatically by. 1-log (median, 366/mL; range, 158 to 756/mL) within 2 to 4 weeks after zoster manifestation and remained at elevated levels for several months. HLA-blocking experiments demonstrated that antiviral CD4 1 T cells were the main IFN-g producers in the first weeks after the onset of zoster (Figure 4). However, we also detected significant VZV reactivity by CD8 1 T cells during this period. This finding demonstrates that even though VZV lysate may be a suboptimal antigen format for detecting VZV-specific CD8 1 T cells, its use in the ELISPOT assay is sufficient for visualizing antiviral CD8 responses early after zoster. In contrast to specific T cells, VZV-IgG antibody levels did not increase upon zoster onset. Only patients 1 and 4 generated a VZV-specific IgM response (Figure 3). Overall, these findings demonstrate that B cell stimulation is inferior to T cell stimulation in response to zoster after allo-hsct. VZV-Specific T Cells Do Not Reconstitute Spontaneously in Nonzoster Patients We next monitored the frequency of VZV-reactive T cells in 12 seropositive patients who did not develop symptomatic VZV reactivation during the first year

4 142 E. Distler et al. Biol Blood Marrow Transplant 14: , 28 A B 5 CD lysate concentration [μg/ml] C CD8 + VZV-infected uninfected IFN-γ spots / 1 5 T cells w/o.5.1 VZV-infected uninfected + IgG2a control + anti-hla-class I + anti-hla-dr IFN-γ spots / 5x1 5 PBMCs 1 2 IFN-γ spots / 5x1 5 PBMCs Figure 1. VZV lysate is recognized by CD4 1 memory T cells. A, PBMCs derived from VZV-seropositive healthy donor 1 were analyzed for reactivity to VZV-infected and uninfected Vero cell lysates in an IFN-g ELISPOTassay. Lysates were added at titrated concentrations to the assay medium. Data are means of duplicate wells and are representative of 6 experiments in 3 different healthy virus carriers. B, CD4 1 and CD8 1 T cells were selected from PBMCs of VZV-seropositive healthy donor 2 and incubated with lysate-loaded autologous dendritic cells in ELISPOT plates. Dendritic cells were induced to mature after the addition of lysate preparations, which is a suitable strategy for promoting lysate-reactive CD8 1 T cell responses [15]. C, ELISPOT reactivity to VZV lysate in PBMCs of seropositive healthy donor 3 was inhibited by monoclonal antibodies with blocking activity against HLA-DR, but not against HLA-class I. Data in panels B and C are means of triplicate wells and are representative of experiments in 5 different healthy immune donors. after undergoing the same allo-hsct protocol [14]. Intraindividual measurements were performed at a median of 3 time points (range, 1 to 3). Interestingly, none of these patients exhibited detectable VZV reactivity over the entire observation period (Figure 5). By analyzing PBMCs of the original donor leukapheresis products, we found that the pretransplantation donor T cell reactivity to VZV did not differ significantly 15 Seropositive Donors Seronegative Donors 125 VZV-reactive T cells / ml PB Figure 2. Quantification of VZV-reactive T cells in healthy immune and nonimmune donors. Diagrams show results of IFN-g ELISPOTreactivity to VZV lysate in PBMCs of 15 seropositive and 5 seronegative healthy donors. The frequency of VZV-reactive T cells/ml of peripheral blood was calculated from the SFC count per plated PBMCs and the PBMC count/ml of peripheral blood. ELISPOT reactivity to PHA did not differ significantly among individual PBMC samples (data not shown). PB, peripheral blood.

5 Biol Blood Marrow Transplant 14: , 28 Zoster drives posttransplant recovery of VZV-specific T cells 1421 VZV-reactive T cells / 5x1 5 PBMCs Stem-cell Donors SD1-DU SD2-MU SD3-MU SD4-MU SD5-MS IgM IgG Patient , Patient months after allo-hsct VZV-reactive T cells / ml PB IgM Patient IgG 2, 1,3 1, Patient 4 Patient , months after allo-hsct 2, 2, 2, 2, VZV-reactive T cells / ml PB Figure 3. Reconstitution of VZV-specific T cell immunity on posttransplantation herpes zoster. Patients 1 to 5 developed zoster during the first year after undergoing T cell depleted allo-hsct. Diagrams show the frequencies of VZV-reactive T cells in leukapheresis-derived PBMCs of stem cell donors before transplantation and in PBMCs of corresponding patients at various time points after transplantation. Results were obtained by IFN-g ELISPOT assay using VZV lysate as the antigen source. Reactivity to uninfected control lysate was not found in any sample. Diagrams also include data of VZVspecific IgM (1, positive; -, negative) and VZV-specific IgG (mu/ml). Arrows indicate time points of zoster onset. SD, stem cell donor; DU, HLA-disparate unrelated; MU, HLA-matched unrelated; MS, HLA-matched sibling; Pat, patient; PB, peripheral blood. between the patients with and without zoster (78 vs 87 SFCs/5 1 5 PBMCs; P 5.25) (Figures 3 and 5). This finding indicates that in the setting of T cell depletion, the VZV-specific T cell frequency in stem cell donors before transplantation was not predictive of future zoster manifestation in allo-hsct recipients, and thus would not be helpful in identifying patients for trials of antiviral prophylaxis. Nevertheless, it remains very difficult to assess the actual contribution of donor T cells to posttransplantation recovery of antiviral T Patient 1 Patient 5 w/o + IgG2a control anti-hla-class I + anti-hla-dr IFN-γ spots / 5x1 5 PBMCs Figure 4. HLA restriction of VZV-reactive T cells during the first weeks after zoster. Shown is IFN-g ELISPOT reactivity to VZV lysate in PBMCs of patients 1 and 5 isolated on days 14 and 35 after zoster onset, respectively. The assay medium contained mabs for blocking HLA-class I or HLA-DR.

6 1422 E. Distler et al. Biol Blood Marrow Transplant 14: , 28 VZV-reactive T cells / 5x1 5 PBMCs Stem-cell Donors SD6-MU SD7-MS SD8-MS SD9-MU SD1-MU SD11-MS SD13-MU SD15-MS Patients w/o Reactivation months after allo-hsct VZV-reactive T cells / ml PB Pat 6 Pat 7 Pat 8 Pat 9 Pat 1 Pat 11 Pat 12 Pat 13 Pat 14 Pat 15 Pat 16 Pat 17 Figure 5. Failure to reconstitute VZV-specific T cells in patients without zoster. Patients 6 to 17 did not develop zoster during the first year after T cell depleted allo-hsct. Diagrams show the frequency of VZV-reactive T cells in leukapheresis-derived PBMCs of stem cell donors before transplantation and in PBMCs of corresponding patients at different time points after transplantation. PBMCs of stem cell donors 12, 14, 16, and 17 were not available. Results were obtained by IFN-g ELISPOT assay. For abbreviations, see Figure 3. cell immunity in patients undergoing T cell depletion in vivo. This is particularly true for recipients of agents that cause severe lymphodepletion, such as alemtuzumab. More studies are needed to determine whether a similar suppression of VZV-specific T cell immunity occurs in patients after T cell replete allo-hsct. We emphasize that our patients received oral antiherpesvirus prophylaxis until day 13 post-hsct, but not thereafter. Whether or not such drugs impede VZV-specific T cell reconstitution by preventing subclinical viremia was not uniformly answered by 2 previous randomized trials [7,9]; however, the consistently low frequency of antiviral T cells in 12 nonzoster patients suggests that avoiding antiviral prophylaxis does not necessarily favor the recovery of VZV-specific T cell immunity. In addition, our results contrast with earlier reports showing that subclinical viremia without progression to herpes zoster already may be sufficient to stimulate VZV-specific T cell proliferation in a significant proportion of allo-hsct recipients [19,2]. Major differences in applied T cell assays (ie, proliferation and cytotoxicity vs ELISPOT), as well as the use of vigorous T cell depletion in our patient cohort, might explain the somewhat contradictory results. Analysis of VZV-Specific Immune Reactivity by Proliferation Assay We decided to also analyze immune reactivity to the VZV-infected Vero cell lysate by proliferation assays in those zoster patients for whom additional PBMC samples were available. We first adapted well-established protocols for proliferation assays for analyzing VZV reactivity [7,9,11-13] to the use of our lysate preparation. Preliminary experiments showed that optimal results could be achieved with an incubation period of 6 days and a VZV lysate concentration of 1 mg/ml (data not shown). The assay medium uninfected VZV-infected 2 cpm x seropos. seroneg. pre post HD HD Pat 1 pre post Pat 3 pre post Pat 4 pre post Pat 5 Figure 6. Detection of VZV reactivity by proliferation assay. PBMCs isolated from patients prior to and after disease manifestation, as well as PBMCs from healthy immune and nonimmune donors (HD), were incubated for 6 days with 1 mg/ml of VZV-infected Vero cell lysate or uninfected control lysate, respectively. Medium and PHA (not shown) served as controls. T cell proliferation was measured by 3 H-thymidine uptake for the last 18 hours of incubation on day 6. Prezoster PBMCs were derived at 2.75 months (patient 1),.75 month (patient 3), 4.5 months (patient 4), and 2.25 months (patient 5) after HSCT. Postzoster PBMCs were obtained at 4.75 months (patient 1), 3 months (patient 3), 6.25 months (patient 4), and 1.75 months (patient 5) after HSCT. For correlation with ELISPOT data, see Figure 3.

7 Biol Blood Marrow Transplant 14: , 28 Zoster drives posttransplant recovery of VZV-specific T cells 1423 proved to be specific since proliferation to VZV lysate was seen in PBMCs of VZV-seropositive healthy donors only, and not in those of a seronegative individual without a history of VZV contact (Figure 6). In addition, proliferation of PBMCs above background values was not found in control wells seeded with the uninfected Vero cell lysate. In subsequent analysis of zoster patients, we detected VZV lysate-specific proliferation in postzoster PBMCs from 3 of 4 patients, which was not evident before disease manifestation (Figure 6). These results are consistent with the IFN-g ELISPOT data obtained from the same PBMC samples (Figure 3). In contrast, patient 4, who failed to respond to the VZV lysate in the proliferation assay, exhibited minor, but still significant, reactivity on the ELISPOT assay. This finding demonstrates the superior sensitivity of the IFN-g ELISPOT assay, allowing the detection of antigenspecific T cells at the single-cell level. Conclusions and Potential Applications Using a newly developed IFN-g ELISPOT assay, we found an early decrease in VZV-reactive T cells in the peripheral blood of VZV-seropositive allo-hsct recipients. Their low numbers and delayed recovery most likely were caused by administration of the lymphocytotoxic antibody alemtuzumab during conditioning therapy [14]. Subsequently, symptomatic VZV reactivation was required to boost antiviral T cells to frequencies comparable to those measured in healthy virus carriers. A similar correlation between posttransplantation recovery of antiviral T cells and virus reactivation has been shown for human cytomegalovirus (CMV) [21,22]. Our study provides some indication of the level of circulating VZV-reactive T cells that may protect against zoster. This conclusion is derived from the finding of frequencies. 1/mL in both immunocompetent donors and patients with zoster. None of the latter developed a second episode of VZV disease during a median period of 14 months (range, 12 to 18 months). Finally, our findings suggest that a vaccination approach, as a substitute for natural resensitization by virus reactivation, may help accelerate the reconstitution of VZV-specific T cell immunity in allo-hsct recipients [13,23]. The use of our assay in prospective trials should help to optimize such vaccination strategies and also help to define a threshold frequency of circulating VZV-reactive T cells that allows for discontinuation of antiviral prophylaxis. ACKNOWLEDGMENTS Financial disclosure: The authors have nothing to disclose. The authors thank Dr C. Meyer, University of Mainz, for providing PBMC samples from the VZV-negative donors. We also thank Michaela Frey and Dr Sandra Kausche for their assistance. REFERENCES 1. Gnann JW Jr, Whitley RJ. Herpes zoster. N Engl J Med. 22; 347: Arvin AM. Cell-mediated immunity to varicella-zoster virus. J Infect Dis. 1992;166:S35-S Locksley RM, Flournoy N, Sullivan KM, et al. Infection with varicella-zoster virus after marrow transplantation. J Infect Dis. 1985;152: Arvin AM. Varicella-zoster virus: pathogenesis, immunity, and clinical management in hematopoietic cell transplant recipients. Biol Blood Marrow Transplant. 2;6: Han CS, Miller W, Haake R, et al. Varicella zoster infection after bone marrow transplantation: incidence, risk factors and complications. Bone Marrow Transplant. 1994;13: Koc Y, Miller KB, Schenkein DP, et al. Varicella zoster virus infections following allogeneic bone marrow transplantation: frequency, risk factors, and clinical outcome. Biol Blood Marrow Transplant. 2;6: Ljungman P, Wilczek H, Gahrton G, et al. Long-term acyclovir prophylaxis in bone marrow transplant recipients and lymphocyte proliferation response to herpes virus antigens in vitro. Bone Marrow Transplant. 1986;1: Thomson KJ, Hart DP, Banerjee L, et al. The effect of low-dose acyclovir on reactivation of varicella zoster virus after allogeneic haematopoietic stem cell transplantation. Bone Marrow Transplant. 25;35: Boeckh M, Kim HW, Flowers MED, et al. Long-term acyclovir for prevention of varicella zoster virus disease after allogeneic hematopoietic cell transplantation: a randomized, double-blind, placebo-controlled study. Blood. 26;17: Erard V, Guthrie KA, Varley C, et al. One-year acyclovir prophylaxis for preventing varicella-zoster virus disease after hematopoietic cell transplantation: no evidence of rebound varicella-zoster virus disease after drug discontinuation. Blood. 27;11: Meyers JD, Flournoy N, Thomas ED. Cell-mediated immunity to varicella-zoster virus after allogeneic marrow transplant. J Infect Dis. 198;141: Kato S, Yabe H, Yabe M, et al. Studies on transfer of varicellazoster virus specific T-cell immunity from bone marrow donor to recipient. Blood. 199;75: Hata A, Asanuma H, Rinki M, et al. Use of an inactivated varicella vaccine in recipients of hematopoietic-cell transplants. N Engl J Med. 22;347: Meyer RG, Britten CM, Wehler D, et al. Prophylactic transfer of CD8-depleted donor lymphocytes after T-cell depleted reduced-intensity transplantation. Blood. 27;19: Herr W, Ranieri E, Olson W, et al. Mature dendritic cells pulsed with freeze-thaw cell lysates define an effective in vitro vaccine designed to elicit EBV-specific CD4 1 and CD8 1 T lymphocyte responses. Blood. 2;96: Frey CR, Sharp MA, Min AS, et al. Identification of CD8 1 T cell epitopes in the immediate early 62 protein (IE62) of varicella-zoster virus, and evaluation of frequency of CD8 1 T cell responses to IE62, by use of IE62 peptides after varicella vaccination. J Infect Dis. 23;188: Jones L, Black AP, Malavige GN, et al. Persistent high frequencies of varicella-zoster virus ORF4 protein specific CD4 1 T cells after primary infection. J Virol. 26;8: Abendroth A, Lin I, Slobedman B, et al. Varicella-zoster virus retains major histocompatibility complex class I proteins in the Golgi compartment of infected cells. J Virol. 21;75:

8 1424 E. Distler et al. Biol Blood Marrow Transplant 14: , Ljungman P, Lönnqvist B, Gahrton G, et al. Clinical and subclinical reactivations of varicella-zoster virus in immunocompromised patients. J Infect Dis. 1986;153: Wilson A, Sharp M, Koropchak CM, et al. Subclinical varicellazoster virus viremia, herpes zoster, and T lymphocyte immunity to varicella-zoster viral antigens after bone marrow transplantation. J Infect Dis. 1992;165: Gratama JW, van Esser JWJ, Lamers CHJ, et al. Tetramerbased quantification of cytomegalovirus (CMV)-specific CD8 1 T lymphocytes in T-cell depleted stem cell grafts and after transplantation may identify patients at risk for progressive CMV infection. Blood. 21;98: Cwynarski K, Ainsworth J, Cobbold M, et al. Direct visualization of cytomegalovirus-specific T-cell reconstitution after allogeneic stem cell transplantation. Blood. 21;97: Redman RL, Nader S, Zerboni L, et al. Early reconstitution of immunity and decreased severity of herpes zoster in bone marrow transplant recipients immunized with inactivated varicella vaccine. J Infect Dis. 1997;176: